The Autonomy-Control Paradox: How Enterprise Business Intelligence Democratization Alters Decision Rights in Corporate Hierarchies

Abstract Digital transformation has reshaped how organizations generate value from data, often through "data democratization" - the widespread distribution of analytical tools across the corporate hierarchy (Vial, 2019). This paper examines a specific instance of that shift: the pharmaceutical sector's move from centralized reporting to interactive business intelligence (BI) and AI-enabled CRM systems, intended to foster data-driven cultures and empower localized decision-making. However, drawing on 15 years of practitioner experience, this paper argues that a contradictory outcome arises: as BI systems become more pervasive, corporate headquarters acquire unprecedented visibility into frontline actions, triggering a recentralization of organizational control. This creates an Autonomy-Control Paradox - a structural condition in which managers are nominally empowered yet simultaneously subjected to granular algorithmic oversight that restricts professional discretion. Drawing on organizational economics (Jensen Orlikowski, 2000), this paper conceptualizes Algorithmic Enclosure in the specific context of enterprise BI governance as a structural condition in which centralized analytical systems expand organizational observability and directive capacity while progressively narrowing independent managerial discretion. The analysis proposes two operational flashpoints - automated targeting versus field agency and data panopticism versus localized sensemaking - and conceptualizes Information System Deviance as a specific form of information-systems workaround behavior in which frontline actors selectively alter, omit, supplement, or bypass system-prescribed data practices to reconcile centralized measurement demands with local operational realities. Building on these dynamics, the paper introduces the Observability-Control Spiral, a recursive enterprise-governance process in which expanded data access increases centralized observability; greater observability enables tighter control; intensified control encourages defensive adaptation; these responses degrade the contextual reliability of information available to the center; and declining confidence in that information generates pressure for further monitoring and control. Eight testable propositions are derived, a governance typology is proposed, and a multi-method empirical roadmap is outlined. The paper contributes to IS research by theorizing how data democratization can generate a recursive governance process through which informational empowerment creates new capacities for centralized control while the resulting control responses progressively weaken the informational basis on which that control depends.